Shallow Foundation Settlement Quantification: Application of Hybridized Adaptive Neuro‐Fuzzy Inference System Model

Abstract

Settlement simulating in cohesion materials is a crucial issue due to complexity of cohesion soil texture. This research emphasis on the implementation of newly developed machine learning models called hybridized Adaptive Neuro-Fuzzy Inference System (ANFIS) with Particle Swarm Optimization (PSO) algorithm, Ant Colony optimizer (ACO), Differential Evolution (DE), and Genetic Algorithm (GA) as efficient approaches to predict settlement of shallow foundation over cohesion soil properties. The width of footing (B), pressure of footing (qa), geometry of footing (L/B), count of SPT blow (N), and ratio of footing embedment (Df/B) are considered as predictive variables. Nonhomogeneity and inconsistency of employed dataset is a major concern during prediction modeling. Hence, two different modeling scenarios (i) preprocessed dataset (PP) and (ii) nonprocessed (initial) dataset (NP) were inspected. To assess the accuracy of the applied hybrid models and standalone one, multiple statistical metrics were computed and analyzed over the training and testing phases. Results indicated ANFIS-PSO model exhibited an accurate and reliable prediction data intelligent and had the highest predictability performance against all employed models. In addition, results demonstrated that data preprocessing is highly essential to be performed prior to building the predictive models. Overall, ANFIS-PSO model showed a robust machine learning for settlement prediction.

Novel hybridized adaptive neuro‐fuzzy inference system models based particle swarm optimization and genetic algorithms for accurate prediction of stress intensity factor

“…Thus, a typical ANFIS model structure has five layers as presented in Figure 1. Therefore, if two input variables in terms of X and Y are supposed, and one considered output variable ( f ), the utilized rules can be described as follows 42 : where A 1 , A 2 , B 1 and B 2 represent the X and Y functions and a 1 , a 2 , b 1 , b 2 , r 1 and r 2 represent the linear output parameters from the first and second rules. Thus, the fives layers that connect ANFIS nodes (Figure 1) can be defined as follows: Layer 1 (Inputs): the nodes are square nodes that create the memberships grade, whereas the input variables ( X and Y ) are translated into linguistic terms using the membership functions as follows: …”

Scouring Depth Assessment Downstream of Weirs Using Hybrid Intelligence Models

“…To evaluate the prediction performance of the hybrid models in forecasting the scouring depth downstream of a weir over both training and testing phases, several performance indices such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Correlation Coefficient (CC) and Wilcox Index (WI) are employed as follows [26,[55][56][57][58][59][60][61]: shows the corresponding mean value of observed and predicted scour depth and N T denotes the dataset number.…”

Displacement Monitoring during the Excavation and Support of Deep Foundation Pit in Complex Environment

“…It can make rational use of the soil's own characteristics, give full play to the mutual bonding between soil and pile-anchor structure, and have the advantages of rapid construction and low cost [11,12]. With the continuous development of urban construction and excavation equipment, the scale of foundation pit engineering is developing towards a larger and deeper direction [13,14]. For the construction process of super high-rise buildings and underground garages, the excavation of the deep foundation pit has little influence on the surrounding buildings and traffic roads, so it is necessary to monitor the dynamic changes of soil indexes around the foundation pit during the excavation process [15][16][17][18].…”